Services

Software Engineer Intern

No sandbox project. You’ll ship something a Fortune 100 client uses, and you’ll be able to name it.

Internship·On-site, HSR Layout, Bangalore, India·Entry level·₹30,000–₹60,000/month

Apply now

About Clientell AI

Clientell AI is a San Francisco company. We run Salesforce and revenue operations for enterprise clients, hands-on, and we build the custom AI agents that live inside those systems to handle the work our clients’ teams don’t have time for. The pod that designs, ships, and stands behind this for our largest accounts sits in HSR Layout, Bangalore.

We’re building that team on purpose: one pod, the same accounts, working together instead of handing tickets to each other. Those accounts include a Fortune 100 pharmaceutical company, a European home improvement retail group, and a US commercial fleet safety manufacturer. We’ll share more account detail as your interview progresses.

We’re seventeen people, and about to be more. The next few hires shape what this company becomes, not just what it ships this quarter.

About this role

Most engineering internships are a holding pen. You get a project nobody depends on, you present it in week sixteen, and it’s deleted in week seventeen. We’re too small for that and honestly we don’t have the spare capacity to run a fake one.

So this is the real thing. You sit with the engineers building AI agents for our client accounts and you build alongside them. Your code goes into what a client uses. That means it gets reviewed properly, it gets tested properly, and if it breaks, someone finds out. That’s the point.

Four to six months, on-site, two seats.

What you’ll actually do

• Ship real code into the agents we run for clients, within your first few weeks. Not a parallel demo repo
• Build the pieces: a tool the agent calls, an integration against a client’s API, an eval harness, an internal dashboard the team actually opens
• Write the evals for what you build. If you can’t measure whether it’s right, you haven’t finished
• Pair with the engineer who owns the account, and get your code reviewed by them line by line
• Read Salesforce or HubSpot API docs yourself and get a call working before you ask anyone
• Take a bug from “something’s weird” to closed, including the part where you find out it wasn’t the bug you thought it was
• Sit in on real client calls, so you hear the problem in the client’s words instead of in a ticket
• Say when something is a bad idea, including when a full-timer suggested it

What makes you a fit

• You can actually build. Python or JavaScript or TypeScript, and you’re comfortable opening a codebase you didn’t write and finding your way around it
• You’ve shipped something someone other than you used. A tool, a bot, a script your college club ran on, a side project with real users. Scale doesn’t matter, the fact that someone else depended on it does
• You’ve wired an LLM into something yourself, and you have a specific story about how it went wrong
• You read the documentation and the error message before you ask. Asking is fine, asking first isn’t
• You finish things. Half the difference between a good intern and a forgettable one is whether the last twenty percent got done
• You can be in the office in HSR Layout, Bangalore, for the full four to six months. This isn’t remote-friendly work
• Realistically this fits someone still in college, just graduated, or very early career. We don’t care about your CGPA and we don’t care where you studied

Use AI like you mean it

You’re going to spend the internship building AI agents. We want to see that instinct already pointed at your own work. If there’s a part of your coursework or your last project an AI could do faster than you, you’ve probably already made it do that part.

Bring one thing you built or automated with AI. Two minutes, screen shared. We’ll ask what it got wrong and how you caught it. “It mostly works” is a worse answer than a specific failure you can describe.

Even better if

• You’ve built evals, tests, or any kind of instrumentation for something non-deterministic
• You can ship a frontend when what you built needs one
• You’ve touched Salesforce, HubSpot, or any CRM, even as a user. We don’t expect it going in
• You’ve contributed to open source, or maintained something long enough to deal with someone else’s bug report
• You’ve won or shipped at a hackathon and the thing survived past the wee…